Effects of Growth Hormone Administration on Bone MineralMetabolism, PTH Sensitivity and PTH Secretory Rhythm in

Postmenopausal Women With Established OsteoporosisFranklin Joseph,1 Aftab M Ahmad,1 Mazhar Ul-Haq,1 Brian H Durham,2 Pauline Whittingham,1 William D Fraser,2

and Jiten P Vora1

ABSTRACT:

Introduction: Growth hormone (GH) replacement improves target organ sensitivity to PTH, PTH circadianrhythm, calcium and phosphate metabolism, bone turnover, and BMD in adult GH-deficient (AGHD) pa-tients. In postmenopausal women with established osteoporosis, GH and insulin like growth factor-1 (IGF-1)concentrations are low, and administration of GH has been shown to increase bone turnover and BMD, butthe mechanisms remain unclear. We studied the effects of GH administration on PTH sensitivity, PTHcircadian rhythm, and bone mineral metabolism in postmenopausal women with established osteoporosis.Materials and Methods: Fourteen postmenopausal women with osteoporosis were compared with 14 healthypremenopausal controls at baseline that then received GH for a period of 12 mo. Patients were hospitalizedfor 24 h before and 1, 3, 6, and 12 mo after GH administration and half-hourly blood and 3-h urine sampleswere collected. PTH, calcium (Ca), phosphate (PO4), nephrogenous cyclic AMP (NcAMP), � C-telopeptideof type 1 collagen (�CTX), procollagen type I amino-terminal propeptide (PINP), and 1,25-dihydroxyvitaminD [1,25(OH)2D] were measured. Circadian rhythm analysis was performed using Chronolab 3.0 and Student’st-test and general linear model ANOVAs for repeated measures were used where appropriate.Results: IGF-1 concentration was significantly lower in the women with established osteoporosis comparedwith controls (101.5 ± 8.9 versus 140.9 ± 10.8 �g/liter; p < 0.05) and increased significantly after 1, 3, 6, and 12mo of GH administration (p < 0.001). Twenty-four-hour mean PTH concentration was higher in the osteo-porotic women (5.4 ± 0.1 pM) than in healthy controls (4.4 ± 0.1 pM, p < 0.001) and decreased after 1 (5.2± 0.1 pM, p < 0.001), 3 (5.0 ± 0.1 pM, p < 0.001), 6 (4.7 ± 0.1 pM, p < 0.001), and 12 mo (4.9 ± 0.1 pM, p < 0.05)of GH administration compared with baseline. NcAMP was significantly lower in osteoporotic women (17.2± 1.2 nM glomerular filtration rate [GFR]) compared with controls (21.4 ± 1.4 nM GFR, p < 0.05) andincreased after 1 (24.2 ± 2.5 nM GFR, p < 0.05), 3 (27.3 ± 1.5 nM GFR, p < 0.001), and 6 mo (32.4 ± 2.5 nMGFR, p < 0.001) compared with baseline. PTH secretion was characterized by two peaks in premenopausalwomen and was altered in postmenopausal women with a sustained increase in PTH concentration. GHadministration also restored a normal PTH secretory pattern in the osteoporotic women. The 24-h meanadjusted serum calcium (ACa) concentration increased at 1 and 3 mo (p < 0.001) and PO4 at 1, 3, 6, and 12mo (p < 0.001). 1,25(OH)2D concentration increased after 3, 6, and 12 mo of GH (p < 0.05). An increase inurine Ca excretion was observed at 3 and 6 mo (p < 0.05), and the renal threshold for maximum tubularphosphate reabsorption rate (TmPO4/GFR) increased after 1, 3, 6, and 12 mo (p < 0.05). �CTX concentrationincreased progressively from 0.74 ± 0.07 �g/liter at baseline to 0.83 ± 0.07 �g/liter (p < 0.05) at 1 mo and 1.07± 0.09 �g/liter (p < 0.01) at 3 mo, with no further increase at 6 or 12 mo. PINP concentration increasedprogressively from baseline (60 ± 5 �g/liter) to 6 mo (126 ± 11 �g/liter, p < 0.001), with no further increase at12 mo. The percentage increase in PINP concentration was significantly higher than �CTX (p < 0.05).Conclusions: Our study shows that GH has a regulatory role in bone mineral metabolism. GH administrationto postmenopausal osteoporotic women improves target organ sensitivity to PTH and bone mineral metabo-lism and alters PTH secretory pattern with greater increases in bone formation than resorption. Thesechanges, resulting in a net positive bone balance, may partly explain the mechanism causing the increase inBMD after long-term administration of GH in postmenopausal women with osteoporosis shown in previousstudies and proposes a further component in the development of age-related postmenopausal osteoporosis.J Bone Miner Res 2008;23:721–729. Published online on November 26, 2007; doi: 10.1359/JBMR.071117

Key words: growth hormone, osteoporosis, PTH, circadian rhythm, �-C-telopeptide of type 1 collagen, pro-collagen type I amino-terminal propeptide, bone turnover, bone markers

The authors state that they have no conflicts of interest.

1Department of Diabetes and Endocrinology, Royal Liverpool University Hospital, Liverpool, United Kingdom; 2Department ofClinical Biochemistry, Royal Liverpool University Hospital, Liverpool, United Kingdom.

JOURNAL OF BONE AND MINERAL RESEARCHVolume 23, Number 5, 2008Published online on November 26, 2007; doi: 10.1359/JBMR.071117© 2008 American Society for Bone and Mineral Research

721

JO705289 721 729 May

INTRODUCTION

BONE LOSS AND the increasing incidence of osteoporosisis an accompaniment of aging. Women undergo two

phases of bone loss—a slow phase with a linear decrease inbone, continuing into old age, and a superimposed, accel-erated transient phase beginning at menopause caused byestrogen deficiency.(1–3) The slow phase in the developmentof osteoporosis has been attributed to alteration of age-related factors resulting in impaired osteoblast function andbone formation. These include growth hormone (GH) andinsulin-like growth factor 1 (IGF-1), both major determi-nants of adult bone mass,(4,5) that decrease with advancingage(6–12) and are lower in women with established osteopo-rosis.(13)

The beneficial effects of GH on bone metabolism andBMD have been shown in, adult GH-deficient (AGHD)patients.(14,15) Target organ insensitivity to the effects ofPTH resulting in increased circulating PTH and abnormalPTH secretion(16) contributes to the development of oste-oporosis in AGHD. GH replacement (GHR) in AGHDpatients has been shown to increase bone and renal PTHreceptor or target cell sensitivity to the effects of PTH andsimultaneously restore PTH secretory rhythm, increasebone turnover markers, 1,25-dihydroxy vitamin D[1,25(OH)2D] concentration, and Ca absorption/reabsorption, thus contributing to the positive effects ofGH on bone.(15)

Postmenopausal women with osteoporosis have high cir-culating PTH concentrations with abnormal PTH circadianrhythm(17,18) and may consequently be insensitive to theeffects of PTH,(19–21) although this has not been shownconclusively. The decline in GH/IGF-1 with aging may con-tribute to these PTH related abnormalities through mecha-nisms similar to that observed in untreated AGHD.(15,16)

GH has been previously administered to healthy elderlywomen and women with postmenopausal osteoporosisand increases in bone turnover and BMD have beenshown.(22–24) However, the mechanisms by which GH ex-erts its beneficial effects on bone in established postmeno-pausal osteoporosis remain unexplained. We thereforestudied the effects of 12 mo of GH administration on PTHsecretory pattern, PTH sensitivity, and bone mineral me-tabolism in postmenopausal women with established oste-oporosis.

MATERIALS AND METHODS

Patients

Women from a community osteoporosis screening pro-gram, with newly diagnosed osteoporosis, were recruited tothe study. All patients had undergone bone densitometricevaluation using a Prodigy Oracle Fan-Beam bone densi-tometer (GE Medical Systems, Giles, Buckinghamshire,UK). T-scores were calculated against a reference popula-tion of UK subjects 20–39 yr of age. Osteoporosis was de-fined according to the WHO criteria with a T-score �−2.5(25) of either the lumbar spine (LS) or femoral neck(FN). Patients with diabetes, ischemic heart disease, heartfailure, renal disease, cancer, chronic illness, vertebral frac-

ture, or any disease or medication such as corticosteroids,affecting bone metabolism were excluded. Subjects wereexcluded if they were receiving hormone replacementtherapy (HRT) or had received HRT in the year beforestart of the study, were on Ca and vitamin D supplements,or had ever been exposed to bisphosphonate therapy.

Fourteen postmenopausal women with osteoporosis(63.4 ± 2.1 [SD] yr; range, 52–79 yr) were recruited. Themean BMD T-score ± SE in the lumbar spine (LS; L2–L4)and femoral neck (FN) was −3.3 ± 0.2 and −2.0 ± 0.2, re-spectively. For baseline comparison 14 healthy premeno-pausal control women (33.9 ± 2.2 yr; range, 25–39 yr) withnormal BMD (LS and FN T-score was 0.3 ± 0.3 and 0.6 ±0.2, respectively) were recruited from a database of volun-teers willing to participate in medical research (Table 1).

Methods

Study visits involved admission to the Metabolic BoneUnit of the Royal Liverpool University Hospital at 1:00p.m. for a period of 25 h. An indwelling venous cannula wasinserted in the antecubital fossa of each patient at the timeof admission, and blood samples were collected every halfhour from 2:00 p.m. on the day of admission to 2:00 p.m. thefollowing day. Samples were centrifuged immediately at−4°C, and serum/plasma was separated to be frozen at−70°C for later analysis. Subjects were provided with 1.5liters of water and encouraged to drink at fairly frequentintervals to maintain hydration and urine samples were col-lected at 3-h intervals between 2:00–11:00 p.m. and 8:00a.m. and 2:00 p.m., and aliquots of these samples werestored at −20°C for later analysis. A 24-h urine volume wasmeasured to estimate fluid balance and no significant vari-ability in the hydration of the individuals was observed.Subjects remained recumbent from 11:00 p.m. to 8:00 a.m.and slept during this period. Each subject was served withstandardized hospital meals at 8:00 a.m., 12:00 p.m., 6:00p.m., and 10:00 p.m. The serving sizes and combinations of

TABLE 1. DEMOGRAPHIC CHARACTERISTICS OF PATIENTS

AND CONTROLS

Osteoporotic women Control subjects

n 14 14Age (yr) 63.4 (2.0)* 33.9 (2.2)Height (m) 1.60 (0.02) 1.63 (0.02)Weight (kg) 63.2 (3.0) 70.7 (3.1)BMI (kg/m2) 24.9 (1.2) 26.8 (1.2)Waist/hip ratio 0.84 (0.02) 0.85 (0.02)Fat mass (kg) 21.6 (2.3) 25.6 (2.5)Fat percentage (%) 33.26 (2.37) 34.32 (2.56)IGF-I (ng/ml) 102.8 (10.0)† 140.9 (10.8)IGF-I SDS −1.21 (0.26) −1.19 (0.28)BMD (T-score)Lumbar spine −3.3 (0.2)‡ 0.3 (0.3)Femoral neck −2.0 (0.2)‡ 0.6 (0.2)

Values are mean (SE).* p < 0.01 osteoporotic women compared with control subjects.† p < 0.05 osteoporotic women compared with control subjects.‡ p < 0.001 osteoporotic women compared with control subjects.

JOSEPH ET AL.722

foods contained recommended daily allowances of all nu-trients including Ca and PO4.

Samples were collected in all controls and subjects withosteoporosis at baseline, after which GH (Humatrope; EliLilly & Co., Basingstoke, Hampshire, UK) was started at astandard daily dose of 0.2 mg, self-injected using an auto-mated pen device (Humatrope-Pen II, Eli Lilly & Co.) at10:00 p.m. every night in the subjects with osteoporosis. GHwas initiated at 0.2 mg/d for 4 wk and titrated by incrementsof 0.1 mg/d every 2 wk, according to IGF-1 concentration.We aimed for a target IGF-1 concentration within ±1 SD ofthe median IGF-1 for a woman 45 yr of age. The IGF-1normal ranges were established with a cohort of 450 healthyadults (age, 18–80 yr; men � 225). The distribution of thevalues obtained was log normal, and consequently, themeasured values were log-transformed before further cal-culations. Means and SD of the log IGF-1 values were cal-culated for age intervals (10 yr). Best fit regression curveswere derived for means and mean minus SD. IGF SDscores (IGF SDS) were calculated by using the formula logIGF-1 – mean/SD. Study visits were repeated 1, 3, 6, and 12mo after the initiation of GH and all patients completed theentire 12-mo study. The local ethics committee approvedthe study, and written informed consent was obtained fromeach patient before recruitment.

Biochemistry

Serum Ca, PO4, creatinine, and albumin were measuredon all samples by standard procedures on an automatedplatform (Hitachi 747; Roche Diagnostics, Lewes, UK). Se-rum Ca was adjusted for albumin.(26) Serum ACa has beenshown to strongly correlate with ionized Ca and has beenfound to be precise in subjects with Ca and albumin withinthe reference range.(26,27) Serum PTH(1-84) was measuredon all samples using the Advantage automated assay plat-form (Nichols Institute, San Juan Capistrano, CA, USA),with a detection limit of 0.5 pM and intra- and interassayCVs of <7% across the working range.

For the 1,25(OH)2D assay, serum was treated with ace-tonitrile and the supernatant purified through C18-OH re-verse phase column to obtain the fraction containing1,25(OH)2D, which after evaporation was measured by ra-dioimmunoassay (IDS, Boldon, UK). Each sample con-tained tritiated 1,25(OH)2D to act as a recovery. The intra-assay CV was <9% and the interassay CV was <12% acrossthe working range, with a detection limit of 15 pM. Serum25-hydroxyvitamin D [25(OH)D] was measured using anRIA kit (DiaSorin, Stillwater,MN) after acetonitrile extrac-tion. The intra-assay CV was <8%, and the interassay CVwas <11% across the working range, with a detection limitof 4 nM.

Serum concentration of type-I collagen-� C-telopeptide(�CTX) and procollagen type-I amino-terminal propeptide(PINP) and osteocalcin were measured on the Elecsys au-tomated platform, which uses electrochemiluminescenceassays (ECLIA; Roche Diagnostics, Lewes, UK). The intra-and interassay CVs for �CTX were <4% and <5%, respec-tively, across the working range, with a detection limit of0.01 �g/liter and the intra- and interassay CVs for PINP,

were <2% and <2.5%, respectively, across the workingrange, with a detection limit of 4 �g/liter. The intra- andinterassay CVs for osteocalcin were both <5% across theworking range with a detection limit of 0.5 �g/L.

Urine creatinine, Ca, and PO4 were analyzed on allsamples using standard laboratory methods (Roche Diag-nostics). Urine values are expressed as molar ratios to cre-atinine (Ca/Cr, PO4/Cr) and as excretion per liter of creat-inine clearance (CCr) by multiplying the urinary ratios andthe serum creatinine to yield the CaE and PO4E, respec-tively. The renal threshold for maximum tubular phosphatereabsorption rate (TmPO4/GFR; mM of GFR) was derivedfrom the nomogram by Walton and Bijvoet.(28) Nephrog-enous cyclic AMP (NcAMP), which is a reliable index ofPTH activity at the level of the kidney,(29) was determinedfrom the formula: NcAMP (nM GFR) � (SCr [�M] ×UcAMP [�M]/Ucr [mM]). Plasma cyclic AMP (PcAMP)was measured by radioimmunoassay (BIOTRAC cAMP;Amersham Pharmacia Biotech, Little Chalfont, UK). Theintra-assay CV was <8%, and interassay CV was <10%across the working range, with a detection limit of 5 nM.Urine cyclic AMP (UcAMP) was measured by in-houseradioimmunoassay (RIA) as previously described.(30) Theintra- and interassay CVs were <8% and 10%, respectively,with a detection limit of 0.2 �M.

IGF-1 was measured with a specific RIA in the presenceof a large excess of IGF-2 (Mediagnost, Tübingen, Ger-many) to block the interference of IGF-binding pro-teins.(31) Intra- and interassay CVs were 1.6% and 6.4%,respectively.

Statistical analysis

Individual and population mean cosinor analysis wasused first to confirm circadian rhythmicity and determinethe circadian rhythm parameters of PTH using CHRONO-LAB 3.0 (Universdade de Vigo, Vigo, Spain), a softwarepackage for analyzing biological time series by least squaresestimation.(15,16,32) The package has previously been wellvalidated and used to analyze PTH and bone marker cir-cadian rhythms in various groups of patients.(15,16,18) Thesoftware thus provides the following circadian parameters:(1) midline estimate statistic of rhythm (MESOR), definedas the rhythm-adjusted mean or the average value of therhythmic function fitted to the data; (2) amplitude, definedas one half the extent of rhythmic change in a cycle ap-proximated by the fitted cosine curve (difference betweenthe maximum and MESOR of the fitted curve); and (3)acrophase, defined as the lag between a defined referencetime (2:00 p.m. of the first day in our study when the fittedperiod is 24 h), and time of peak value of the crest time inthe cosine curve fitted to the data. A p value for the rejec-tion of the zero-amplitude (no rhythm) assumption is alsodetermined for each individual series and for the group.The method used by the program allows analysis of hybriddata (time series sampled from a group of subjects, eachrepresented by an individual series).(33,34) Bingham’s test,developed for testing cosinor parameters and part ofCHRONOLAB 3.0 software, was used to determine thesignificance of the differences of cosinor-derived circadianrhythm parameters between subjects.

EFFECTS OF GH ADMINISTRATION IN ESTABLISHED OSTEOPOROSIS 723

After the confirmation of concerted circadian rhythmsfurther analysis of the more extensively studied PTHrhythm was performed as the next step. Over and above thediurnal variation, the PTH rhythm has previously beenshown to have bimodal peaks in healthy individuals (earlyevening and nocturnal) and previous studies in pathologicalconditions have shown alterations during the time periodsof these peaks.(15,16,18,33,35) Based on these previous data,time points for further analysis were selected from the in-dividual peaks and. the time of onset was defined as thetime of first occurrence of at least three consecutivesamples exceeding the mean levels of PTH obtained be-tween 8:00 a.m. and 2:00 p.m. by >1 SD.(36)

General linear model ANOVA (GLM ANOVA) for re-peated measures was used to analyze the data. Repeated-measures ANOVA assumes normally distributed errors,equal variances, and sphericity. The Kolmogorov-Smirnovtest was used to confirm normal distribution and Levenes’test for equality of variances. Mauchly’s test indicated thatthe sphericity assumption was violated and degrees of free-dom were corrected using Greenhouse-Geisser estimates ofsphericity (� � 0.53). This method has been validated forsimilar comparisons.(16,37) For all analyses, p < 0.05 wasconsidered significant. Values are expressed as the mean ±SE unless otherwise stated.

RESULTS

GH dose and IGF-I levels

IGF-I concentration (Fig. 1A) was significantly lower inthe women with osteoporosis compared with the controls(101.5 ± 8.9 versus 140.9 ± 10.8 �g/liter; p < 0.05). In thetreated patients, the mean GH dose (Fig. 1B) was 0.2 ± 0.01mg/d at 1 mo and increasing to 0.39 ± 0.01 mg/d at 3 mo(p < 0.001) and further titrated to 0.56 ± 0.04 mg/d at 6 mo(p < 0.001 compared with baseline; p < 0.01 compared with3 mo). Maximum dose was achieved at 6 mo in all patients,and further increases were not tolerated, and three patientshad dose reductions from 6 to 12 mo because of increasedmusculo-skeletal pains. The mean GH dose was 0.49 ± 0.05mg/d at 12 mo (p < 0.001 compared with baseline, p < 0.05compared with 3 mo, and p � not significant [NS] com-pared with 6 mo). Mean serum IGF-I increased significantlyfrom 101.5 ± 8.9 to 128.3 ± 12.1 �g/liter by 1 mo (p < 0.001),157.2 ± 15.8 �g/liter at 3 mo (p < 0.001), 172.9 ± 13.6 �g/literat 6 mo (p < 0.001, compared with baseline; p � 0.06,

compared with 3 mo), and 166.9 ± 14.8 �g/liter at 12 mo(p < 0.001, compared with baseline; p � NS, compared with3 and 6 mo). Similarly IGF-I SDS increased from −1.26 ±0.27 at baseline to −0.62 ± 0.29 at 1 mo (p < 0.001), −0.03 ±0.26 at 3 mo (p < 0.001), 0.28 ± 0.26 at 6 mo (p < 0.001,compared with baseline; p � 0.07, compared with 3 mo),and 0.19 ± 0.26 at 12 mo (p < 0.001, compared with baseline;p � NS, compared with 3 and 6 mo).

PTH

Twenty-four-hour mean PTH concentration (Fig. 2A)was higher in the osteoporotic women (5.4 ± 0.1 pM) thanin healthy controls (4.4 ± 0.1 pM, p < 0.001). After GHadministration, 24-h mean PTH concentration decreasedprogressively from baseline (5.4 ± 0.1 pM) to 1 (5.2 ± 0.1pM, p < 0.001), 3 (5.0 ± 0.1 pM, p < 0.001), and 6 mo (4.7 ±0.1 pM, p < 0.001), with maximum reduction in PTH con-centration at 6 mo. PTH concentrations increased signifi-cantly by 12 mo (4.9 ± 0.1 pM, p < 0.05 compared with 6mo) but remained below baseline concentrations.

Individual and population cosinor analyses for circulatingPTH (Figs. 3 and 4) showed significant circadian rhythmsfor all healthy controls and all osteoporotic patients at allvisits (p < 0.001) but with differences between patients andcontrols and changes after GH administration. The meanPTH MESOR was significantly higher in the osteoporoticwomen than in the controls (5.4 ± 0.3 versus 4.4 ± 0.3 pM,p � 0.03), but there was no significant difference in theamplitude (0.7 ± 0.1 versus 0.5 ± 0.1 pM for osteoporoticwomen and controls, respectively, p � 0.22) or acrophase(−156 ± 16° versus −178 ± 16° for osteoporotic women andcontrols, respectively, p � 0.36). After GH administrationto the osteoporotic women, mean PTH MESOR decreasedby 3 mo (5.0 ± 0.3 pM, p < 0.05), with a further decrease at6 mo (4.5 ± 0.2 pM, p < 0.001) when maximum reduction inmean PTH MESOR was observed. PTH MESOR rose sig-nificantly by 12 mo (4.9 ± 0.1 pM, p < 0.05 compared with6 mo) but remained below baseline (5.3 ± 0.3 pM, p < 0.05).The amplitude and acrophase of the PTH circadian rhythmdid not change significantly after GH administration. A re-duction in mean percentage increase(33) in PTH concentra-tion between 2:00 and 11:00 p.m., without significant changein the maximum percentage increase(33) indicated a nar-rower afternoon/evening peak, following 3, 6, and 12 mo ofGH administration. The maximum percentage increase andthe mean percentage change in PTH concentration be-

FIG. 1. GH doses and IGF-1 concentra-tions.

JOSEPH ET AL.724

tween 11:30 p.m. and 8:00 a.m.(33) was significantly lower inthe osteoporotic women as compared with the controls rep-resenting a less marked nocturnal peak. The maximum andmean percentage change in PTH concentration overnightincreased significantly after GH administration indicatingrestoration of the nocturnal peak.

NcAMP

NcAMP (Fig. 2B) was significantly lower in osteoporoticwomen (17.2 ± 1.2 nM GFR) compared with controls (21.4± 1.4 nM GFR, p < 0.05). NcAMP increased after 1 mo ofGH administration (24.2 ± 2.5 nM GFR, p < 0.05) andremained elevated at 3 (27.3 ± 1.5 nM GFR, p < 0.001) and6 mo (32.4 ± 2.5 nM GFR, p < 0.001) compared with base-line (17.2 ± 1.2 nM GFR) and returned to levels not sig-nificantly different to baseline at 12 mo (14.8 ± 1.6 nMGFR).

Calcium

Twenty-four-hour mean ACa concentration (Fig. 2C)was not significantly different in the osteoporotic women(2.36 ± 0.004 mM, p < 0.05) compared with controls (2.35 ±0.004 mM). The 24-h mean ACa concentration increasedprogressively after 1 (2.40 ± 0.002 mM, p < 0.001) and 3 mo(2.38 ± 0.004 mM, p < 0.001) of GH compared with baseline(2.36 ± 0.004 mM) but returned to concentrations that were

FIG. 2. Biochemical parameters in osteo-porotic women compared with controls andbefore and after GH administration. NS*, nosignificant difference, osteoporotic womencompared with control subjects; *p < 0.05 os-teoporotic women compared with controlsubjects; **p < 0.01 osteoporotic womencompared with control subjects; ***p < 0.001osteoporotic women compared with controlsubjects. NS+, no significant difference com-pared with baseline; +p < 0.05 compared withbaseline; ++p < 0.01 compared with baseline;+++p < 0.001 compared with baseline. NS€, nosignificant difference compared with 1 mo; €p< 0.05 compared with 1 mo; €€p < 0.01 com-pared with 1 mo; €€€p < 0.001 compared with1 mo. NS£No significant difference comparedwith 3 mo; £p < 0.05 compared with 3 mo; ££p< 0.01 compared with 3 mo; £££p < 0.001 com-pared with 3 mo. NS�No significant differencecompared with 6 mo; �p < 0.05 comparedwith 6 mo; ��p < 0.01 compared with 6 mo;���p < 0.001 compared with 6 mo.

FIG. 3. Cosinor-derived PTH rhythms for control subjects andosteoporotic women. Uninterrupted lines represent osteoporoticwomen and dotted lines represent control subjects. Arrows (un-interrupted, osteoporotic women; dotted, control subjects) repre-sent the acrophase in degrees.

EFFECTS OF GH ADMINISTRATION IN ESTABLISHED OSTEOPOROSIS 725

not significantly different from baseline at 6 (2.35 ± 0.002mM) and 12 mo (2.34 ± 0.002 mM).

Phosphate

Twenty-four-hour mean PO4 concentration (Fig. 2E) waslower in osteoporotic women (1.11 ± 0.01 mM, p < 0.05)compared with controls (1.15 ± 0.01 mM). Twenty-four-hour mean PO4 concentration increased progressively after1 (1.18 ± 0.01 mM), 3 (1.23 ± 0.01 mM), and 6 mo (1.27 ±0.01 mM) and remained elevated with no further increase at12 mo (1.28 ± 0.01 mM) of GH administration comparedwith baseline (1.11 ± 0.01 mM, p < 0.001).

Vitamin D metabolites

No significant difference in serum 25(OH)D3 concentra-tion was observed between controls (42.0 ± 6.1 nM) andosteoporotic women (47.6 ± 6.1 nM, p � 0.53). 25(OH)D3

concentrations were not significantly different after 1 (45.1± 4.2 nM), 3 (49.9 ± 5.3 nM), and 12 (42.2 ± 4.8 nM) mo ofGH administration. An increase was observed at 6 mo (54.5± 4.5 nM, p < 0.05) compared with baseline (47.6 ± 3.6 nM).No significant difference in serum 1,25(OH)2D (Fig. 2D)concentration was observed between controls (74.0 ± 8.2pM) and osteoporotic women (78.1 ± 8.2 pM, p � 0.73; Fig.2D). 1,25(OH)2D concentrations increased by 3 mo (99.4 ±10.0 pM, p < 0.001) and were maintained at 6 (95.7 ± 9.5pM, p < 0.05) and 12 mo (99.9 ± 11.6 pM, p < 0.01; Fig. 2D).

Urine calcium excretion

Ca/Cr (0.6 ± 0.03 versus 0.4 ± 0.03; p < 0.001) and CaE(Fig. 2F; 0.05 ± 0.002 versus 0.03 ± 0.002 mM CCr; p <

0.001) were higher in the osteoporotic women comparedwith controls. Ca/Cr increased after 3 (0.8 ± 0.04; p < 0.001)and 6 (0.9 ± 0.04; p < 0.001) mo of GH and was not signifi-cantly different at 1 (0.7 ± 0.04; p � 0.1) and 12 mo (0.7 ±0.04; p � 0.2) compared with baseline (0.6 ± 0.03). CaE alsoincreased similarly (baseline, 0.05 ± 0.002 mM CCr; 1 mo,0.06 ± 0.003 mM CCr, p � NS; 3 mo 0.07 ± 0.003 mM CCr,p < 0.001; 6 mo 0.08 ± 0.003 mM CCr, p < 0.001; 12 mo 0.06± 0.003 mM CCr, p � NS).

Urine phosphate excretion and TmPO4/GFR

PO4/Cr (2.2 ± 0.12 versus 2.5 ± 0.12; p � 0.1) and PO4E(Fig. 2H; 0.18 ± 0.009 mM CCr versus 0.19 ± 0.009 mM CCr;p � 0.3) were not significantly different in the two groups.PO4/Cr increased significantly after 1 (2.8 ± 0.15; p < 0.001),3 (3.2 ± 0.16; p < 0.001), 6 (3.3 ± 0.17; p < 0.001), and 12 mo(2.7 ± 0.13; p < 0.05) of GH compared with baseline (2.2 ±0.12) with similar increases in PO4E (baseline, 0.18 ± 0.009mM CCr; 1 mo, 0.22 ± 0.012 mM CCr, p < 0.001; 3 mo, 0.25± 0.012 mM CCr, p < 0.001; 6 mo, 0.28 ± 0.014 mM CCr,p < 0.001; 12 mo, 0.21 ± 0.012 mM CCr, p < 0.05).

TmPO4/GFR (Fig. 2G) was not significantly different be-tween controls (0.98 ± 0.02 mM GFR) and osteoporoticwomen (0.96 ± 0.02 mM GFR, p � 0.6). TmPO4/GFR in-creased after GH administration for 1 mo (0.99 ± 0.02 mMGFR, p < 0.05) and remained elevated at 3 (0.99 ± 0.02 mMGFR, p < 0.05), 6 (1.01 ± 0.02 mM GFR, p < 0.001), and 12(1.08 ± 0.02 mM GFR, p < 0.001) mo compared with base-line.

Markers of bone turnover

�CTX concentrations (Fig. 2I) were significantly higherin osteoporotic women (0.74 ± 0.07�g/liter) compared withcontrols (0.20 ± 0.07 �g/liter, p < 0.001). After GH admin-istration �CTX concentrations increased progressivelyfrom baseline (0.74 ± 0.07�g/liter) to 1 (0.83 ± 0.07 �g/liter,p < 0.05) and 3 mo (1.07 ± 0.09 �g/liter, p < 0.001), with nofurther increase seen at 6 (1.18 ± 0.10 �g/liter, p < 0.001compared with baseline) and 12 mo (1.08 ± 0.12 �g/liter,p < 0.001 compared with baseline).

PINP (Fig. 2J) concentrations were significantly higher inosteoporotic women (60 ± 5 �g/liter) compared with con-trols (35 ± 5 �g/liter, p < 0.01). PINP concentrations in-creased progressively from baseline (60 ± 5 �g/liter) to 1 (69± 5 �g/liter, p < 0.001), 3 (93 ± 6 �g/liter, p < 0.001), and 6mo (126 ± 11 �g/liter, p < 0.001). The increase was main-tained after 12 mo (122 ± 14 �g/liter, p < 0.001) of GHadministration.

Osteocalcin concentrations increased progressively frombaseline (34 ± 3 �g/liter) to 1 (37 ± 3 �g/liter, p < 0.05), 3 (48± 3 �g/liter, p < 0.01), and 6 mo (61 ± 4 �g/liter, p < 0.001).The increase was maintained after 12 mo (59 ± 5 �g/liter,p < 0.001 compared with baseline) of GH administration.

The percentage increase in PINP concentration was sig-nificantly higher than �CTX after 6 (76 ± 25% versus 142 ±25%, p < 0.05) and 12 mo (61 ± 25% versus 133 ± 25%,p < 0.05) of GH administration and the percentage increasein osteocalcin concentration was significantly >12 mo (61 ±25% versus 76 ± 25%, p < 0.05).

FIG. 4. Cosinor-derived PTH rhythms before and 6 mo afterGH administration. Uninterrupted lines represent osteoporoticwomen before GH administration and dotted lines represent os-teoporotic women after 6 mo of GH administration. Arrows (un-interrupted, osteoporotic women before GH administration; dot-ted, osteoporotic women after 6 mo of GH administration)represent the acrophase in degrees.

JOSEPH ET AL.726

DISCUSSION

Postmenopausal women with osteoporosis have lowercirculating IGF-1 concentration with higher 24-h meanPTH and lower NcAMP concentration compared withhealthy premenopausal women with normal BMD. GH ad-ministration resulted in increased IGF-1 concentration, de-creased PTH concentration, and increased NcAMP. GHadministration also resulted in an increase in 1,25(OH)2D,serum ACa, serum PO4, TmPO4/GFR, and biochemicalmarkers of bone turnover with a greater percentage in-crease in markers of bone formation than resorption. Ourfindings indicate a decrease in target organ sensitivity toPTH in postmenopausal women with osteoporosis that in-creased after GH administration. GH administration re-stored the circadian rhythm of PTH that was altered inosteoporotic women.

Changes in PTH and bone metabolism in postmeno-pausal women with osteoporosis(19–21) have previouslybeen mainly attributed to oestrogen deficiency after themenopause, diminished response to vitamin D and to agingitself.(38–45) GH effects on PTH in osteoporosis have onlybeen studied in the short term and with the co-administration of other calcitropic agents with variable re-sults including decreased,(22) unchanged,(22,46) or increasedconcentration(47) attributed to increased bone turnover andenhanced mobilization of skeletal Ca or increased Ca ab-sorption.(46) Previous studies have not measured NcAMP,which reflects the activity of PTH in both physiological andpathophysiological states and is a reliable index of PTHfunction.(48) Because NcAMP excretion parallels changesin PTH secretion,(49) the reciprocal increase in NcAMP ex-cretion with decreasing PTH concentration, observed in ourstudy indicates increased renal sensitivity to the effects ofPTH after GH administration. The observed decrease inNcAMP back to baseline levels at 12 mo may be a reflec-tion of the mean GH dose decrease at 12 mo. However, thesignificantly lower PTH concentration at 12 mo comparedwith baseline in the presence of a NcAMP concentrationsimilar to baseline still suggests an improvement in PTHsensitivity having achieved a new equilibrium. PTH showeda sustained increase between 2:00 and 11:00 p.m. with areduced nocturnal rise in osteoporotic women as previouslyshown.(18) After GH administration the PTH secretory pat-tern changed significantly with restoration to a rhythm re-sembling that similar to healthy control subjects,(17,49,50)

supporting a role for GH in regulating PTH secretoryrhythm.

1,25(OH)2D and Ca absorption decrease with aging andthe normal increase in 1,25(OH)2D in response to infusionsof PTH is blunted possibly because of decreased renal25(OH)D 1�-hydroxylase sensitivity to PTH.(51) It has alsobeen suggested that women with osteoporosis have a defectin renal calcium conservation(52) and the higher urine Caexcretion in association with the relatively higher PTH con-centration in our subjects suggests renal resistance to theeffects of PTH may contribute to this. Although the con-centrations of 25(OH)D3 were similar in both groups stud-ied, they were in the low normal range, and this may haveaffected PTH secretion. The initial increase in circulating

Ca with no change in urine Ca excretion after GH admin-istration may partly reflect increased renal sensitivity toPTH resulting in renal Ca reabsorption as a direct effect ofPTH.(53) A possible increase in 1-� hydroxylase activityresulting in increased 1,25(OH)2D production,(46,54,55) andsubsequent Ca absorption may also have contributed.(56)

The simultaneous increase in urine Ca excretion probablyreflects renal regulation of the higher filtered Ca load. By12 mo, however, the circulating Ca and renal Ca excretionachieve equilibrium possibly with increased Ca utilizationfor bone matrix formation. PO4 levels have been shown toincrease or remain unchanged in aging postmenopausalwomen with a negative relationship between PO4 and vita-min D. In our osteoporotic subjects with normal vitamin Dconcentrations, we found a lower circulating PO4 concen-tration associated with a marginally lower TmPO4/GFRcompared with controls, although the difference was notstatistically significant. An initial increase in both theTmPO4/GFR and serum PO4 concentration was observedafter GH administration, possibly as a result of a directanti-phosphaturic effect of GH/IGF-1(57,58) and increased1,25(OH)2D3 activity.(59,60) Urine PO4 excretion also in-creased in parallel with increasing serum PO4 and thereforea higher filtered PO4 load before a new equilibrium wasestablished at 12 mo when circulating phosphate andTmPO4/GFR reached a plateau and phosphate excretionbegan to decrease. Increased phosphaturia could also be areflection of the change in renal sensitivity to PTH duringGH therapy, but clearly the dominant effect of these hor-mones on PO4 is varying with time.

GH administration to osteoporotic patients simulta-neously increases markers of both bone formation and re-sorption and thus no increase in BMD is seen in the shortterm but BMD increases after prolonged GH administra-tion.(22) Our data confirm a simultaneous increase in boneresorption and formation with the increase in bone forma-tion markers becoming significantly higher than resorptiononly by 6 mo, possibly explaining the delay in increase inBMD after GH administration. The sequence of changes inbone turnover markers is different from the response toexogenously administered fixed dose PTH, which results inan early increase in bone formation markers preceding anyincrease in resorption by about a month. The apparent dif-ference may be a result of several individual factors or morelikely a combination of these factors. This may include thegradual increase in GH dose, the subsequent gradual in-crease in circulating GH/IGF-1 concentrations, paracrineand autocrine effects of IGF-1 and other growth factors inthe bone microenvironment, an increase in bone cell sensi-tivity to endogenous PTH, and changes in bone cell respon-siveness to changes in the circadian rhythm of endogenousPTH.

In conclusion, our results suggest that lower GH andIGF-1 concentration are associated with target organ insen-sitivity to the effects of PTH and abnormal PTH circadianrhythm in postmenopausal women with osteoporosis.These differences may be a result of an age-related processalone, with osteoporosis being an epi-phenomenon or maybe the result of a combination of these two processes. GHadministration restores sensitivity to PTH and PTH rhythm

EFFECTS OF GH ADMINISTRATION IN ESTABLISHED OSTEOPOROSIS 727

with subsequent changes in bone turnover, Ca, and PO4

metabolism resulting in positive bone balance contributingto the delayed increase in BMD shown in previous studies.We believe our study consolidates previous studies and pro-poses a further component in the development of osteopo-rosis in postmenopausal women.

REFERENCES

1. Riggs BL, O’Fallon WM, Muhs J, O’Connor MK, Kumar R,Melton L Jr 1998 Long-term effects of calcium supplementa-tion on serum parathyroid hormone level, bone turnover, andbone loss in elderly women. J Bone Miner Res 13:168–174.

2. Marcus R 1991 Skeletal aging. Undeerstanding the functionaland structural basis of osteoporosis. Trends Endocrinol Metab2:53–58.

3. Krolner B, Pors Nielsen S 1982 Bone mineral content of thelumbar spine in normal and osteoporotic women: Cross-sectional and longitudinal studies. Clin Sci (Lond) 62:329–336.

4. Boonen S, Aerssens J, Broos P, Pelemans W, Dequeker J 1995Age-related bone loss and senile osteoporosis: Evidence forboth secondary hyperparathyroidism and skeletal growth fac-tor deficiency in the elderly. Aging (Milano) 7:414–422.

5. Boonen S, Lesaffre E, Aerssens J, Pelemans W, Dequeker J,Bouillon R 1996 Deficiency of the growth hormone-insulin-like growth factor-I axis potentially involved in age-related al-terations in body composition. Gerontology 42:330–338.

6. Donahue L, Hunter S, Sherblom A, Rosen C 1990 Age-relatedchanges in serum insulin-like growth factor-binding proteins inwomen. J Clin Endocrinol Metab 71:575–579.

7. Finkelstein JS, Roffwarg H, Boyar P, Kream J, Hellman L 1972Age-related changes in the twenty-four hour spontaneous se-cretion of growth hormone in normal individuals. J Clin En-docrinol Metab 35:665–670.

8. Zadik Z, Chalew SA, Mc Carter RJ Jr, Meistas M, KowarskiAA 1985 The influence of age on the 24- hour integrated con-centration of growth hormone in normal individuals. J ClinEndocrinol Metab 60:513–516.

9. Ho KY, Evans WS, Blizzard RM, Veldhuis JD, Merriam GR,Samojlik E, Furlanetto R, Rogol AD, Kaiser DL, Thorner MO1987 Effects of sex and age on the 24-hour profile of growthhormone secretion in man: Importance of endogenous estra-diol concentrations. J Clin Endocrinol Metab 64:51–58.

10. Corpas E, Harman SM, Piñeyro MA, Roberson R, BlackmanMR 1992 Growth hormone (GH)-releasing hormone-(1-29)twice daily reverses the decreased GH and insulin-like growthfactor-I levels in old men. J Clin Endocrinol Metab 75:530–535.

11. Vermeulen A 1987 Nyctohemeral growth hormone profiles inyoung and aged men: Correlation with somatomedin-C levels.J Clin Endocrinol Metab 64:884–888.

12. Iranmanesh A, Lizarralde G, Veldhuis JD 1991 Age and rela-tive adiposity are specific negative determinants of the fre-quency and amplitude of growth hormone (GH) secretorybursts and the half-life of endogenous GH in healthy men. JClin Endocrinol Metab 73:1081–1088.

13. Landin-Wilhelmsen K, Wilhelmsen L, Bengtsson B 1999 Post-menopausal osteoporosis is more related to hormonal aberra-tions than to lifestyle factors. Clin Endocrinol (Oxf) 51:387–394.

14. Johannsson G, Rosén T, Bosaeus I, Sjöström L, Bengtsson BA1996 Two years of growth hormone (GH) treatment increasesbone mineral content and density in hypopituitary patientswith adult-onset GH deficiency. J Clin Endocrinol Metab81:2865–2873.

15. Ahmad AM, Thomas J, Clewes A, Hopkins MT, Guzder R,Ibrahim H, Durham BH, Vora JP, Fraser WD 2003 Effects ofgrowth hormone replacement on parathyroid hormone sensi-tivity and bone mineral metabolism. J Clin Endocrinol Metab88:2860–2868.

16. Ahmad AM, Hopkins MT, Fraser WD, Ooi CG, Durham BH,Vora JP 2003 Parathyroid hormone secretory pattern, circulat-

ing activity, and effect on bone turnover in adult growth hor-mone deficiency. Bone 32:170–179.

17. Eastell R, Calvo MS, Burritt MF, Offord KP, Russell RG,Riggs BL 1992 Abnormalities in circadian patterns of boneresorption and renal calcium conservation in type I osteopo-rosis. J Clin Endocrinol Metab 74:487–494.

18. Fraser WD, Logue FC, Christie JP, Gallacher SJ, Cameron D,O’Reilly DS, Beastall GH, Boyle IT 1998 Alteration of thecircadian rhythm of intact parathyroid hormone and serumphosphate in women with established postmenopausal osteo-porosis. Osteoporos Int 8:121–126.

19. Cosman F, Shen V, Herrington BS 1990 Mechanism of estro-gen action in osteoporosis treatment as assessed by human(1-34) PTH infusion. In: Christiansen C, Overgaard K (eds.)Osteoporosis III. Osteopress, Copenhagen, Denmark, pp. 976–978.

20. Lindsay R, Sweeney A 1976 Urinary cyclic-AMP in osteopo-rosis. Scott Med J 21:231.

21. Cosman F, Shen V, Herrington B, Lindsay R 1991 Response ofthe parathyroid gland to infusion of human parathyroid hor-mone-(1-34) [PTH-(1-34)]: Demonstration of suppression ofendogenous secretion using immunoradiometric intact PTH-(1-84) assay. J Clin Endocrinol Metab 73:1345–1351.

22. Landin-Wilhelmsen K, Nilsson A, Bosaeus I, Bengtsson B 2003Growth hormone increases bone mineral content in postmeno-pausal osteoporosis: A randomized placebo-controlled trial. JBone Miner Res 18:393–405.

23. Kassem M, Brixen K, Blum WF, Mosekilde L, Eriksen EF 1994Normal osteoclastic and osteoblastic responses to exogenousgrowth hormone in patients with postmenopausal spinal oste-oporosis. J Bone Miner Res 9:1365–1370.

24. Holloway L, Butterfield G, Hintz RL, Gesundheit N, Marcus R1994 Effects of recombinant human growth hormone on meta-bolic indices, body composition, and bone turnover in healthyelderly women. J Clin Endocrinol Metab 79:470–479.

25. WHO Study Group 1994 Assessment of fracture risk and itsapplication to screening for postmenopausal osteoporosis. Re-port of a WHO Study Group. World Health Organ Tech RepSer 843:1–129.

26. Gardner MD, Dryburgh FJ, Fyffe JA, Jenkins AS 1981 Pre-dictive value of derived calcium figures based on the measure-ment of ionised calcium. Ann Clin Biochem 18:106–109.

27. White TF, Farndon JR, Conceicao SC, Laker MF, Ward MK,Kerr DN 1986 Serum calcium status in health and disease: Acomparison of measured and derived parameters. Clin ChimActa 157:199–213.

28. Walton RJ, Bijvoet OL 1975 Nomogram for derivation of renalthreshold phosphate concentration. Lancet 2:309–310.

29. Drezner MK, Neelon FA, Curtis HB, Lebovitz HE 1976 Renalcyclic adenosine monophosphate: An accurate index of para-thyroid function. Metabolism 25:1103–1112.

30. O’Reilly DS, Fraser WD, Penney MD, Logue FC, Cowan RA,Williams BC, Walters G 1986 Arginine infusion blocks theaction of parathyroid hormone but not arginine vasopressin onthe renal tubule in man. J Endocrinol 111:501–506.

31. Blum WF, Brier BH 1994 Radioimmunoassays for IGF’s andIGFBP’s. Growth Regul 4:11–19.

32. Mojon A, Fernandez JR, Hermida RC 1992 Chronolab: Aninteractive software package for chronobilogic time seriesanalysis written for the Macintosh computer. Chronobiol Int9:403–412.

33. Joseph F, Chan BY, Durham BH, Ahmad AM, Vinjamuri S,Gallagher JA, Vora JP, Fraser WD 2007 The circadian rhythmof osteoprotegerin and its association with parathyroid hor-mone secretion. J Clin Endocrinol Metab 92:3230–3238.

34. Fernandez JR, Hermida RC 1998 Inferential statistical methodfor analysis of nonsinusoidal hybrid time series with unequi-distant observations. Chronobiol Int 15:191–204.

35. White HD, Ahmad AM, Durham BH, Patwala A, Whitting-ham P, Fraser WD, Vora JP 2005 Growth hormone replace-ment is important for the restoration of parathyroid hormone

JOSEPH ET AL.728

sensitivity and improvement in bone metabolism in older adultgrowth hormone-deficient patients. J Clin Endocrinol Metab90:3371–3380.

36. Luboshitzky R, Shen-Orr Z, Herer P 2003 Middle-aged mensecrete less testosterone at night than young healthy men. JClin Endocrinol Metab 88:3160–3166.

37. White HD, Ahmad AM, Syed AA, Clewes A, Peter R, VoraJP, Fraser WD 2004 Gender variation in PTH sensitivity andrhythmicity following growth hormone replacement in adultgrowth hormone-deficient patients. Clin Endocrinol (Oxf)60:516–526.

38. Marcus R, Madvig P, Young G 1984 Age-related changes inparathyroid hormone and parathyroid hormone action in nor-mal humans. J Clin Endocrinol Metab 58:223–230.

39. Khosla S, Atkinson EJ, Melton L Jr, Riggs BL 1997 Effects ofage and estrogen status on serum parathyroid hormone levelsand biochemical markers of bone turnover in women: A popu-lation-based study. J Clin Endocrinol Metab 82:1522–1527.

40. Imanaka S, Onishi T, Morimoto S, Takamoto S, Kohno H,Kumahara Y 1985 Comparison of renal responses to synthetichuman PTH(1-34) administration in normal young and elderlymale subjects. Calcif Tissue Int 37:357–362.

41. Epstein S, Bryce G, Hinman JW, Miller ON, Riggs BL, Hui SL,Johnston CC 1986 The influence of age on bone mineral regu-lating hormones. Bone 7:421–425.

42. Orwoll ES, Meier DE 1986 Alterations in calcium, vitamin Dand parathyroid hormone physiology in normal men with ag-ing: Relationship to the development of senile osteoporosis. JClin Endocrinol Metab 63:1262–1269.

43. Minisola S, Pacitti MT, Scarda A, Rosso R, Romagnoli E,Carnevale V, Scarnecchia L, Mazzuoli GF 1993 Serum ionizedcalcium, parathyroid hormone and related variables: Effect ofage and sex. Bone Miner 23:183–193.

44. Ledger GA, Burritt MF, Kao PC, O’Fallon WM, Riggs BL,Khosla S 1994 Abnormalities of parathyroid hormone secre-tion in elderly women that are reversible by short term therapywith 1,25-dihydroxyvitamin D3. J Clin Endocrinol Metab79:211–216.

45. Buchanan JR, Myers CA, Greer RB III 1988 Effect of declin-ing renal function on bone density in aging women. Calcif Tis-sue Int 43:1–6.

46. Lieberman S, Holloway L, Marcus R, Hoffman A 1994 Inter-actions of growth hormone and parathyroid hormone in renalphosphate, calcium, and calcitriol metabolism and bone re-modeling in postmenopausal women. J Bone Miner Res9:1723–1728.

47. Marcus R, Butterfield G, Holloway L, Gilliland L, Baylink DJ,Hintz RL, Sherman BM 1990 Effects of short term adminis-tration of recombinant human growth hormone to elderlypeople. J Clin Endocrinol Metab 70:519–527.

48. Broadus AE, Mahaffey JE, Bartter FC, Neer RM 1977 Ne-phrogenous cyclic adenosine monophosphate as a parathyroidfunction test. J Clin Invest 60:771–783.

49. Logue F, Fraser W, O’Reilly D, Beastall G 1989 The circadian

rhythm of intact parathyroid hormone (1-84) and nephrog-enous cyclic adenosine monophosphate in normal men. J En-docrinol 121:R1–R3.

50. el-Hajj Fuleihan G, Klerman EB, Brown EN, Choe Y, BrownEM, Czeisler CA 1997 The parathyroid hormone circadianrhythm is truly endogenous–a general clinical research centerstudy. J Clin Endocrinol Metab 82:281–286.

51. Tsai KS, Heath H, Kumar R, Riggs BL 1984 Impaired vitaminD metabolism with aging in women. Possible role in pathogen-esis of senile osteoporosis. J Clin Invest 73:1668–1672.

52. Heshmati HM, Khosla S, Burritt MF, O’Fallon WM, Riggs BL1998 A defect in renal calcium conservation may contribute tothe pathogenesis of postmenopausal osteoporosis. J Clin En-docrinol Metab 83:1916–1920.

53. Juppner H, Brown EM, Kronenberg HM 1999 Parathyroidhormone. In: Favus M (ed.) Primer on the Metabolic BoneDiseases and Disorders of Mineral Metabolism, 4th ed. Lip-pincott Williams & Wilkins, Philadelphia, PA, USA, pp. 80–87.

54. Spencer EM, Tobiassen O 1981 The mechanism of the actionof growth hormone on vitamin D metabolism in the rat. En-docrinology 108:1064–1070.

55. Spanos E, Barrett D, MacIntyre I, Pike JW, Safilian EF, Haus-sler MR 1978 Effect of growth hormone on vitamin D metabo-lism. Nature 273:246–247.

56. Chipman JJ, Zerwekh J, Nicar M, Marks J, Pak CY 1980 Effectof growth hormone administration: Reciprocal changes in se-rum 1 alpha,25-dihydroxyvitamin D and intestinal calcium ab-sorption. J Clin Endocrinol Metab 51:321–324.

57. Kruse K, Kracht U, Gopfert G 1982 Renal threshold phos-phate concentration (TmPO4/GFR). Arch Dis Child 57:217–223.

58. Caversazio J, Bonjour JP, Fleisch H 1989 Adaptation of tubu-lar phosphate transport: Relation between phosphate require-ment as influenced by growth and supply. Adv Exp Med Biol128:107–111.

59. Ogle GD, Rosenberg AR, Kainer G 1992 Renal effects ofgrowth hormone. II. Electrolyte homeostasis and body com-position. Paediatr Nephrol 6:483–489.

60. Bouillon R 1991 Growth hormone and bone. Horm Res36(Suppl 1):49–55.

Address reprint requests to:Franklin Joseph, MBBS, MRCP

Department of Diabetes and EndocrinologyLink 7C

Royal Liverpool University HospitalPrescot Street

Liverpool L78XP, UKE-mail: drfrankjoseph@yahoo.co.uk

Received in original form May 21, 2007; revised form October 10,2007; accepted November 21, 2007.

EFFECTS OF GH ADMINISTRATION IN ESTABLISHED OSTEOPOROSIS 729